INDEX

Explanations

names of people

This neuron detects personal names (proper names of people).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 пу

0.26

При

0.25

 detectable

0.24

 ти

0.24

 wastage

0.24

ВА

0.23

 discernible

0.23

По

0.23

Для

0.23

ОР

0.23

POSITIVE LOGITS

 Williams

0.34

 coordinadora

0.32

 Singh

0.31

 William

0.31

 Hernandez

0.31

 Hernández

0.30

 Michael

0.30

 Robert

0.30

 Kumar

0.30

 Martin

0.30

Activations Density 0.584%